This code reproduces the figures and statistics presented in "Putting the Significance of Spectral Peaks on the Level: Implications for the 1470-Yr Peak in Greenland d18O", published in the Journal of Climate (2022).  Code is written for Matlab (R2021b) and requires the econometrics toolbox.  Below is a brief description of the contents.

Files:

1. fig123.m: Reproduces figures 1, 2, and 3 in the main text.
2. fig4.m:     Reproduces figure 4 in the main text.
3. fig_supplemental.m: An additional figure showing random realization according with (a) H0(1,0), (c) H0(1,1), (e) H0(3,2), and (g) H0(5,5) in black lines and the compliment for H1 in red lines.  Panels at right show spectral estimates made from the time series at left with the vertical black dashed line at 1/(1470 yr).
4. reported_stats.m: Calculation of many of the relevant statistics reported in the main manuscript.
5. all_arma.m: code for computing all ARMA(p,q) models ranging from p equal to 1 to 5 and q equal 0 to 5.
6. GISP2.m, GISPsynth.mat, and EX_GISP.mat: data, synthetic data, and supporting statistics for rendering figure 4.

Folders:

Toolbox: Supporting code for spectral leveling.  See the header file for each script for a more detailed description and all_arma.m for an example of how to call the toolbox.


Peter Huybers
Harvard, 2022
phuybers@fas.harvard.edu
